Polymyositis and dermatomyositis are uncommon disease of auto- immune etiology. Most patients have disease responsive to prednisone, however, approximately 20% of PM/DM patient either do not respond to steriods, respond incompletely, or have intolerable adverse effects of therapy. In these individuals, cytotoxic drugs such as azathioprine or methotrexate are often given. The rate of adverse reactions to cytotoxic agents is high, however, and a significant proportion of patients will remain nonresponsive or intolerant of therapy. A number of anecdotal reports describe improvement in symptoms of PM/DM in response wither plasma exchange or leukapheresis. In the majority of these cases, therapy with steroids and/or cytotoxic agents was given simultaneously, and the effect of apheresis could not be analyzed apart from that of other therapies. The purpose of this study was to define the role of apheresis in the treatment if PM/DM in a randomized, sham-controlled, double-blind fashion. Patients non-responsive or intolerant of high-dose prednisone were randomized to either: (1) plasma exchange, (2) leukapheresis, or (3) sham apheresis. The apheresis procedures were performed 3 times weekly for 4 weeks. The patients and the investigators assessing response to therapy were kept blinded as to the procedures performed. Patients who did not improve with apheresis were subsequently given cyclophosphamide 1.0 g/M2 IV monthly for 6 months. Thirty-two patients have entered the study; analysis of responses in muscle strength, CPK, aldolase, LDH, SGOT, and muscle biopsy performed after entry of the first 22 patients revealed that differences among groups had not reached statistical significance. Response rates will be reanalyzed after accrual of 45 patients.